Ankur Singla 2539b6a984 [DistributedInference] Relax the assertion for uniqueness of blob name across external inputs and outputs (#72492) ... Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/72492 Having same blob name present in external inputs and external outputs is a valid case, so relaxing the validation for that. Reviewed By: yyetim Differential Revision: D34062055 fbshipit-source-id: 6772ef9c3259da221207d14e5cc93a7777002ef2 (cherry picked from commit 0de66a2941cee8847a4689a4eb0091dfff82dd85)		2022-02-09 21:50:47 +00:00
..
custom	use irange for loops (#70248)	2022-01-06 23:14:29 -08:00
nql	use irange for loops (#70248)	2022-01-06 23:14:29 -08:00
annotations.cc
annotations.h
backend_cutting.cc
backend_cutting.h
backend_cutting_test.cc
backend_transformer_base.cc
backend_transformer_base.h
bound_shape_inference_test.cc
bound_shape_inferencer.cc
bound_shape_inferencer.h
CMakeLists.txt
converter.cc
converter.h
converter_nomigraph_test.cc
dead_code_elim.cc
dead_code_elim_test.cc
device.cc
device.h
device_test.cc
distributed.cc	[DistributedInference] Relax the assertion for uniqueness of blob name across external inputs and outputs (#72492)	2022-02-09 21:50:47 +00:00
distributed.h
distributed_converter.cc
distributed_test.cc	[DistributedInference] Relax the assertion for uniqueness of blob name across external inputs and outputs (#72492)	2022-02-09 21:50:47 +00:00
fakefp16_transform.cc
fakefp16_transform.h
fusion.cc
fusion.h
glow_net_transform.cc
glow_net_transform.h
mobile.cc
mobile.h
mobile_test.cc
onnx_convert.h
onnxifi_op.cc
onnxifi_op.h	use irange for loops (#70248)	2022-01-06 23:14:29 -08:00
onnxifi_transformer.cc
onnxifi_transformer.h
optimize_ideep.cc
optimize_ideep.h
optimizer.cc
optimizer.h
passes.cc
passes.h
shape_info.cc
shape_info.h
split_slss_test.cc
tvm_transformer.cc
tvm_transformer.h